Currently patents last about 17 years, and copyright about 100 years. Originally both were about 14 years, equal to two 7-year apprentice terms. ((See my post Where did the patent term come from?; also C. Michael White, “Why a Seventeen Year Patent,” 38 J. Pat. Off. Soc’y 839 (1956) (describing the historical basis for the seventeen-year patent term).)) As noted by Michele Boldrin & David Levine,
In 1976, the term of copyright, which since 1909 had been 28 years plus a renewal term of 28 years, was increased to the life of the author plus 50 years. … Since the Sony Bono Copyright Term Extension Act of 1998, copyright protection in the U.S. is life of the author plus 70 years, or 90 years for works without an author. If we take the remaining life of an author to be roughly 35 years, this would mean 105 years of protection. ((Michele Boldrin & David K. Levine, “Market Size and Intellectual Property Protection,” International Economic Review, Vol. 50, Issue 3, pp. 855-881 (August 2009).))
There have been attempts to estimate what the “optimal” terms should be, assuming patent and copyright do promote some innovation and creativity, and based on mainstreamish utilitarian assumptions. They are all much shorter than current terms–yet the longer terms still persist (thanks, Mickey Mouse). For example, Rufus Pollock, in Forever Minus a Day? Some Theory and Empirics of Optimal Copyright, concludes that a 15 year copyright term is “optimal”–i.e., about 85 years shorter than the present system. Boldrin and Levine, the authors of Against Intellectual Monopoly, performed a detailed, careful econometric analysis in 2009 and concluded that the “optimal” copyright term ought to be about 2 years, and patents about 5 to 10 years. ((Boldrin & Levine, “Market Size and Intellectual Property Protection.”)) I say “optimal” in scare quotes because Boldrin and Levine’s conclusion is based on the assumption that patent and copyright are beneficial–an assumption Boldrin and Levine reject, so that their actual view is that the optimal term should be zero–that is, that we ought to abolish patent and copyright. As Levine explained in an email to me about this paper (my emphasis added):
These results assume that without IP a smaller fraction of the social surplus is recovered by the inventor/creator than with IP–that is the sense in which we assume that copyright and patent have some benevolent effect. That is, we assume that IP actually increases innovation. What we ignore is the downstream effect of IP–that future innovation/creation is reduced due to existing rights. We also ignore the fact that the producers who count–the marginal ones–are not especially likely to benefit from IP. In practice I suspect that those two are the main reasons that empirically IP just doesn’t increase creation/innovation. Of course our calculation ignores any other social costs of IP other than the standard economic loss due to monopoly power.
And yes: I think the optimal IP term is zero. I wouldn’t argue that a perfectly benevolent government (however improbable that might be) with really really good information (adding improbability to improbability) couldn’t make some small social improvement by occasionally awarding a monopoly for some particularly important and costly invention/creation. But leaving those improbabilities aside, the critical problem is if you concede an inch the lobbyists will soon occupy the entire tent. The only safe thing is an out and out clear cut prohibition on IP.
Other studies also indicate a need for shorter IP terms. In “Why a Seventeen Year Patent,” 38 J. Pat. Off. Soc’y 839 (1956), C. Michael White describes the historical basis for the seventeen-year patent term and proposes shortened terms. Merges & Nelson, in “On the Complex Economics of Patent Scope,” pp. 868–70, argue that most economic models of patent scope and duration focus on the relation between breadth, duration, and incentives to innovate, without giving serious consideration to the social costs of greater duration and breadth in the form of retarded subsequent improvement. ((See also Jonathan Barnett, Cultivating the Genetic Commons, observing that “There is little determinative empirical evidence to settle theoretical speculation over the optimal scope and duration of patent protection.” Citing D.J. Wright, “Optimal patent breadth and length with costly imitation,” 17 Intl. J. Industrial Org. 419, 426 (1999).))
The bottom line is that even arguments with pro-IP assumptions imply that the patent term should be cut in half and the copyright term slashed by at least 85 years (Amazon CEO Jeff Bezos favors reducing some patent terms to 3–5 years). ((An Open Letter From Jeff Bezos On The Subject Of Patents (March 2000).)) This is one reason I have argued that any serious IP reform should include first and foremost a significant reduction in the patent and copyright terms. ((See my article Reducing the Cost of IP Law.)) Unfortunately, IP reform efforts never reduce patent and copyright terms, which is one reason I have argued that these reform efforts are not serious or radical at all. ((See my articles and posts ; Patent Reform is Here! O Joy!; Prior User Rights and Patent Reform; “Reducing the Cost of IP Law“; “Radical Patent Reform Is Not on the Way.”))